SN 1997BS IN M66: ANOTHER EXTRAGALACTIC η CARINAE ANALOG? 1. and

Transcription

1 Preprint typeset using L A TEX style emulateapj SN 1997BS IN M66: ANOTHER EXTRAGALACTIC η CARINAE ANALOG? 1 Schuyler D. Van Dyk IPAC/Caltech, Mailcode , Pasadena CA Chien Y. Peng Steward Observatory, Astronomy Department, University of Arizona, Tucson, AZ Jennifer Y. King, Alexei V. Filippenko, Richard R. Treffers, Weidong Li Astronomy Department, University of California, Berkeley, CA and Michael W. Richmond Physics Department, Rochester Institute of Technology, Rochester, NY To appear in the PASP (2000 Dec issue) ABSTRACT We report on SN 1997bs in NGC 3627 (M66), the first supernova discovered by the Lick Observatory Supernova Search using the 0.75-m Katzman Automatic Imaging Telescope (KAIT). Based on its earlytime optical spectrum, SN 1997bs was classified as Type IIn. However, from the BV RI light curves obtained by KAIT early in the supernova s evolution, and F555W and F814W light curves obtained from Hubble Space Telescope archival WFPC2 images at late times, we question the identification of SN 1997bs as a bona fide supernova. We believe that it is more likely a super-outburst of a very massive luminous blue variable star, analogous to η Carinae, and similar to SN 1961V in NGC 1058 (Filippenko et al [AJ, 110, 2261]) and SN 1954J ( Variable 12 ) in NGC 2403 (Humphreys & Davidson 1994 [PASP, 106, 1025]). The progenitor may have survived the outburst, since the SN is seen in early 1998 at m F555W =23.4, about 0.5 mag fainter than the progenitor identified by Van Dyk et al. (1999, [AJ, 118, 2331]) in a pre-discovery image. Based on analysis of its environment in the Hubble Space Telescope images, the progenitor was not in an H II region or association of massive stars. The recent discovery of additional objects with properties similar to those of SN 1997bs suggests that the heterogeneous class of Type IIn supernovae consists in part of impostors. Subject headings: supernovae: general supernovae: individual (SN 1997bs) stars: evolution stars: variables: other galaxies: individual (NGC 3627) galaxies: stellar content 1. INTRODUCTION The ultimate fate of the most massive stars in our Galaxy and other galaxies is still not well understood. The post-main-sequence evolution of stars with masses M > M, theoretically, should take them possibly through the red supergiant phase, or directly to the luminous blue variable (LBV) phase, en route to becoming hydrogen-deficient or hydrogenless Wolf-Rayet stars toward the end of their lives (e.g., Langer et al. 1994; Stothers & Chin 1996). Although the number of known examples of LBVs is small (Humphreys & Davidson 1994), the spectacular cases of η Carinae (e.g., Davidson & Humphreys 1997) and the Pistol Star (Figer et al. 1998, 1999) show that they go through spectacular eruptive phases of mass ejection. It is assumed that very massive stars ultimately explode as supernovae (SNe; e.g., Woosley, Langer, & Weaver 1993), but it is still possible that they fail to become supernovae and core collapse directly to form black holes (MacFadyen & Woosley 1999). The spectroscopic characteristics of SNe have been recently reviewed by Filippenko (1997, 2000a). In brief, Type I SNe (SNe I), lacking hydrogen in their optical spectra, divide into the classical SNe Ia, which probably arise from exploding white dwarf stars, and SNe Ib/c, which appear to be more closely related to Type II SNe. All SNe II show hydrogen in their optical spectra and are thought to arise from the explosions of massive stars (M > 8 M ). Great variation exists in the strength and profile of the hydrogen lines of SNe II. A growing number of peculiar SNe II show a hydrogen emission-line combination of a narrow component atop a broader component (sometimes having a complex shape), with no P-Cygni-like absorption trough. These are the SNe II- narrow (SNe IIn; Schlegel 1990; Filippenko 1991). As of 2000 May, 43 SNe IIn have been identified (from the online Asiago SN Catalog 2 and the IAU Circulars). The Balmer emission has been interpreted as arising from interaction of the SN ultraviolet radiation and the SN ejecta with a dense circumstellar medium set up by the progenitor in a pre-sn phase. Evidence is accumulating that SNe IIn arise from very massive stars (e.g., Weiler, Panagia, & Sramek 1990; Sollerman, Cumming, & Lundqvist 1998). It seems likely that the old Zwicky (1965) Type V SNe, examples of which include SNe 1954J and 1961V, and which are now generally considered 1 BASED IN PART ON OBSERVATIONS MADE WITH THE NASA/ESA HUBBLE SPACE TELESCOPE, OBTAINED FROM THE DATA ARCHIVE OF THE SPACE TELESCOPE SCIENCE INSTITUTE, WHICH IS OPERATED BY THE ASSOCIATION OF UNIVER- SITIES FOR RESEARCH IN ASTRONOMY, INC., UNDER NASA CONTRACT NAS supern/snean.txt 1

2 2 to have been peculiar SNe II (Doggett & Branch 1985; Filippenko 1991), are related to SNe IIn. For instance, SN 1986J, a prototypical SN IIn, was originally classified as Type V (Rupen et al. 1987). However, particularly in the case of SN 1961V (Goodrich et al. 1989; Filippenko et al. 1995), it appears that not all of the Zwicky types were genuine SNe, strictly defined to be the violent destruction of a star at the end of its life. By extension, then, it is intriguingly possible that not all SNe IIn are real SNe. Some SNe IIn could represent a pre-sn cataclysmic event of a very massive star. This contamination of the SN IIn subtype complicates calculations of the production rate of neutron stars and the chemical evolution of galaxies, and adds a new wrinkle in the evolution of the most massive stars. In this paper, we discuss the example of the SN IIn 1997bs in NGC 3627 (M66). SN 1997bs was the first SN discovered by the Lick Observatory Supernova Search (LOSS 3, the direct successor of the former search at Leuschner Observatory), which is robotically conducted with the 0.75-m Katzman Automatic Imaging Telescope (KAIT; see Filippenko et al. 2000). Treffers et al. (1997) reported the SN discovery, from an image (Figure 1) obtained on 1997 April 15 (UT dates are used throughout this paper). The SN brightness at discovery was roughly estimated to be R 17 mag. An image from April 10 did not show anything at the location of SN 1997bs, to a limiting magnitude R The position of the SN is RA=11 h 20 m , Dec= (J2000.0; Cavagna 1997), which is about 11. 2westand69. 9 south of the galaxy s nucleus. NGC 3627 has been host to two other known SNe, the Type II SN 1973R (Ciatti & Rosino 1977) and the Type Ia SN 1989B (Wells et al. 1994). Here we present evidence that SN 1997bs was not a genuine SN, through spectroscopic and photometric observations made at Lick and also photometry from archival Hubble Space Telescope (HST) images. 2. OBSERVATIONS 2.1. Lick Spectroscopy On 1997 April 16, CCD spectra of SN 1997bs were obtained by Filippenko, Barth, & Gilbert (1997) with the Kast spectrograph (Miller & Stone 1993) on the Shane 3- m reflector at Lick Observatory. The slit width was 2,and three different gratings and grisms were used to span the wavelength range Å with a typical resolution of about 7 Å. The spectra were extracted and calibrated following standard procedures, and we took special care to remove telluric absorption features through division by the spectrum of the flux comparison star. Owing to variable seeing, the absolute flux calibration is only approximate, but the shape of the spectrum should be reasonably accurate. The spectrum (Figure 2) is dominated by relatively narrow (FWHM 1000 km s 1 ) Balmer emission lines on a featureless continuum. Many weaker Fe II emission lines are also evident. From this spectrum, Filippenko et al. (1997) classified SN 1997bs as Type IIn KAIT Photometry BV RI photometry was obtained with KAIT over about 60 days following discovery before it was no longer detectable. Images of NGC 3627 in each of the four bands were obtained on 1994 January 13 at the KPNO 2.1-m telescope by L. A. Wells (who kindly provided them to us). These pre-sn images served as templates which were subtracted from the SN images in each of the four bands (see Filippenko et al. 1986; Richmond et al. 1995). First, for each band, the template was registered to a SN image, so that stars in the field were aligned, and the template image was convolved with a Gaussian to match the FWHM of stars in the SN image. Then, the stars in the template image were flux-scaled to match the stars in the SN image, and, finally, the template was subtracted from the SN image, leaving behind a residual image containing only the SN. This procedure was followed for each observational epoch. Photometry of the SN in each band and for each epoch was done on the template-subtracted image using the APPHOT package in IRAF 4.Measurements of the field stars 3 and 4 (see Wells et al. 1994) were also obtained from the original (not subtracted) images in each band for each epoch, this time determining the sky background from an annulus around each field star. In addition, BV RI images of NGC 3627 were obtained on 1999 December 17 with the 1.5-m telescope at Palomar Observatory, under photometric conditions. In Table 1 we list the magnitudes in each band for stars 3 and 4 derived from the Palomar observations. The agreement with the values given in Table 2 of Wells et al. (1994) for these stars is quite good. With our magnitudes for these stars we then transformed the KAIT photometry of SN 1997bs onto the Johnson-Cousins system HST Archival Image Photometry Deep HST WFPC2 images of NGC 3627 in the bands F555W and F814W were obtained from the image archive. These images were made in pairs (to facilitate removal of cosmic rays) between late 1997 and early 1998 during several epochs. They were originally intended to measure the distance to the galaxy using Cepheid variables (GO program 6549), with the goal of determining the absolute magnitude of SN 1989B (Saha et al. 1999; their Table 1 lists the exposures). Saha et al. identify SN 1997bs as variable star C2-V23, yet they did not recognize it as the SN and note only that C2-V23 and others are definitely variable... and...may be transient variables such as novae. A cosmic-ray split pair of WFPC2 F606W images made on 1994 December 28 was also obtained from the archive. Van Dyk et al. (1999; their Figure 7) display this image pair, as well as representative F555W images showing the SN. To measure the brightness of SN 1997bs in the WFPC2 images, we employed point-spread function (PSF) fitting photometry performed by DAOPHOT (Stetson 1987) and ALLSTAR within IRAF. We used the Tiny Tim routine (Krist 1995) to produce an artificial point-spread function (PSF) image, simulated to have stellar spectral type A, im- 3 bait/kait.html 4 IRAF (Image Reduction and Analysis Facility) is distributed by the National Optical Astronomy Observatories, which are operated by the Association of Universities for Research in Astronomy, Inc., under cooperative agreement with the National Science Foundation.

3 3 age radius 0. 5, and full width at half maximum of The artificial PSF was created at the pixel position of the SN in the archive images. Since the stellar profiles in the images were actually somewhat broader than the artificial PSF, the latter was broadened using a Gaussian to match the former. Instrumental photometry was aperture corrected, using measurements of artificial stars of known magnitude placed randomly in the field, and converted to magnitudes using the photometric synthetic zeropoints in Holtzman et al. (1995; their Table 9), appropriate for a 0. 5 aperture radius. A correction for the gain ratio to the 7 e ADU 1 gain state was applied. 3. RESULTS In Table 2 we list the KAIT photometry for SN 1997bs, while Table 3 gives the photometry derived from the archival HST F555W and F814W images. Our HST photometry of SN 1997bs agrees quite well with that from Saha et al. (1999; their Table 3), to within 0.02 mag at F555W and 0.04 mag at F814W. In Figure 3 we show the combined BV RI light curves for the SN, including both the KAIT and HST photometry. The F555W and F814W measurements correspond approximately to V and I magnitudes, respectively. These WFPC2 synthetic magnitudes were not converted to the Johnson-Cousins system following (for example) the transformations given by Holtzman et al. (1995); for an emission-line object such as SN 1997bs, this conversion may not be entirely correct or appropriate. SN 1997bs was apparently discovered before maximum brightness in R, and generally at or near maximum in the other bands. The SN light curves decline from peak brightness over the next 20 days by 1 mag. The decline in V is less drastic after about JD In R and I, the light curves appear to reach a plateau of roughly constant magnitude in each band. (The observational coverage in B terminates after JD ) The KAIT photometry in all bands ends on JD , after a span of nearly 60 days of monitoring. There is a long, 150-day gap in our coverage before we are able to extend the light curves using the HST photometry in (approximately) V and I bands. It is apparent in V that the light curve began a much steeper decline on or before JD ; the slope of the decline is quite similar to the slope of the earlier post-maximum decline. The V light curve, again, achieves a more gradual decline after about JD In fact, after JD , the V magnitude of the SN appears to reach another plateau, at 23.4 mag. The I light curve at late times appears to gradually decline, at a slower rate than the earlier post-maximum decline (the SN is partially saturated in the first-epoch F814W image, so we can only give a lower limit to its brightness). In Figure 4 we show the color evolution of the SN. For the first 60 days, the SN color becomes generally redder after discovery. This is best exemplified by the evolution of the V I color, which near maximum already is at V I 0.62 mag, but, even before the end of KAIT monitoring, has reached V I 1.43 mag. By the time monitoring resumes, with the inclusion of the HST data, the SN is very red, at V I 3.4 mag. The red color of the SN is obvious in the color composite image shown by Saha et al. (1999; their Figure 1). The SN possibly may have been much redder than this prior to its recovery in the HST images: it becomes bluer at late times, to V I 3 mag in the last F555W/F814W image pair. 4. DISCUSSION Although the spectrum of SN 1997bs indicates that it is a SN IIn, its photometric nature is unusual, both in the behavior of the light curves and in absolute magnitude. Since the distance modulus of NGC 3627 has been measured with Cepheids (µ = mag; Saha et al. 1999), we can adjust our observed light curves of SN 1997bs to an absolute scale. We would like to correct the absolute light curves for extinction, but the amount of reddening toward SN 1997bs is not known. The Galactic foreground reddening toward NGC 3627 is E(B V )=0.03 mag (Schlegel et al. 1998). However, SN 1997bs occurred at the edge of a dust lane in NGC 3627, so it is likely that reddening internal to the galaxy at the site of the SN is appreciable and spatially variable. We have attempted to estimate the reddening, based on the colors of stars in the SN environment (see below), yet the stellar colors, particularly for the bluest stars, are consistent with the foreground reddening estimate. From the light curves, we find that the B V color near maximum was 0.67 mag. If the intrinsic color was close to B V =0.0, which is what one might expect for SNe II, then E(B V ) 0.7 mag. However, the spectrum of SN 1997bs dereddened by this amount would make the SN continuum unphysically blue, so already we know the reddening cannot be this large. The emission-line intensity ratios of Hα/Hβ/Hγ are 2.6/1.0/0.5, which, if anything, are slightly flatter than what one expects for Case B recombination (2.8/1.0/0.47); so, there is no evidence for reddening of the emission lines. We measure an equivalent width of 1.8 Å for the narrow Na I D absorption line at the redshift of the SN (unfortunately, the two line components are not resolved). Though with low-resolution spectra one cannot confidently translate this into a reddening (one actually needs high-dispersion spectra), an analysis similar to that of Filippenko, Porter, & Sargent (1990) suggests a reddening of E(B V )=0.38 mag. Moreover, when this type of analysis was applied to the SN Ic 1987M by Jeffery et al. (1991), they argued that it overestimates E(B V ) by a factor 1.8. If we scale our estimate for SN 1997bs by this factor, then E(B V )=0.21 mag. We assume this value to be the total (but quite uncertain) reddening to the SN and correct the observed light curves for it, using a standard Galactic reddening law (Cardelli, Clayton, & Mathis 1989), A V =3.1E(B V ). SN 1997bs appears to be unusual with respect to most other SNe II (or other SN types), due to its very faint absolute magnitude: maximum for the SN was at only M V 13.8 mag, about 4 to 4.5 mag fainter than a typical SN II. In Figure 5 we show the object s unusual photometric evolution through a comparison of the absolute V light curves of SN 1997bs and of the SNe IIn 1988Z (Turatto et al. 1993), 1987B (Schlegel et al. 1996), and 1994Y (Ho et al. 2000). (Distances to the latter three SNe were calculated from their recession velocities, obtained from the NASA/IPAC Extragalactic Database and an assumed dis-

4 4 tance scale H 0 =65kms 1 Mpc 1. No extinction corrections have been applied for these SNe.) The light curves of the three comparison SNe IIn have been scaled, so that their maximum observed brightness occurs at approximately the same epoch as SN 1997bs. One can see that, besides the obvious underluminosity, the post-maximum decline of SN 1997bs is very different in behavior from that of the other SNe. One also notices from Figure 5 that, at late times, the F555W light curve (in particular) appears to reach a plateau once again, at 7.5 mag(or 23.4 mag from Figure 3). None of the other SNe IIn, particularly SN 1994Y, show this sort of behavior. Van Dyk et al. (1999) found in the F606W image from 1994 December 28, about 27.5 months before the discovery of SN 1997bs, a star at the exact position of the SN, with m F606W =22.86 ± 0.16 mag. Assuming m F606W V, the star had an absolute magnitude M V 8.1 mag, at the distance of NGC 3627 and corrected for our assumed reddening. Van Dyk et al. associated this star with the progenitor of SN 1997bs. The late-time brightness of SN 1997bs is 0.5 mag fainter than the brightness of this precursor star (see Figure 6). However, the unusual apparent flattening-out of the latetime light curve suggests that the star may have survived the explosion. This raises the question of whether SN 1997bs was really a bona fide supernova, or instead was something altogether different. We believe that the evidence presented in this paper indicates that SN 1997bs was more likely the superoutburst of a very massive LBV star, possibly analogous to the enormous eruptions experienced by η Carinae (Davidson & Humphreys 1997). The V light curve, in particular, shows that SN 1997bs brightened by nearly 6 mag before fading, possibly through the formation of optically thick dust, given how red it became. SN 1997bs is not alone in its unusual characteristics. A similar object may be the peculiar SN IIn 1961V in NGC SN 1961V was originally classified as Type V by Zwicky (1964, 1965) and had probably the most bizarre light curve ever recorded for a SN. Its progenitor was identified as a very luminous star, Mpg 0 12 mag, visible in many photographs of NGC 1058 prior to the explosion (Bertola 1964; Zwicky 1964; Klemola 1986). Goodrich et al. (1989) and Filippenko et al. (1995) identified the possible survivor of the SN 1961V event, the latter study via HST WFPC imaging, and both studies conclude that perhaps SN 1961V was not a genuine supernova, but rather also the super-outburst of a luminous blue variable, similar to η Car. We can compare the photometric behavior of SN 1997bs with that of the η Car-like variables discussed thoroughly by Humphreys, Davidson, & Smith (1999). In Figure 6 we show the light curve of SN 1997bs, along with the schematic light curves for η Car, SN 1961V, and SN 1954J (also known as Variable 12, V12, in NGC 2403; Humphreys & Davidson 1994). (Interestingly, both η Car and SN 1954J were originally classified as Type V SNe by Zwicky 1964 and, e.g., Schaefer 1996, respectively.) V -band light curves are shown for SN 1997bs, η Car (Humphreys et al. 1999), and SN 1954J (Tammann & Sandage 1968), while the light curve for SN 1961V is photographic (Humphreys et al. 1999). We have adjusted all of the curves to absolute magnitudes: the distances to SN 1961V (8 Mpc), SN 1954J (3.2 Mpc), and η Car (2300 pc) are from Goodrich et al. (1989), Freedman & Madore (1988), and Davidson & Humphreys (1997), respectively. (The curve for SN 1997bs is that from Figure 5, with the addition of the 1994 pre-discovery observation; the curves for the other three objects have not been extinction-corrected.) All of the curves have been further adjusted, so that the maxima occur at the same time. Overall, there are a number of similarities between these various light curves, although there are also some notable differences. The eruption of η Car lasted 20 years, while the outburst of SN 1997bs, with its shorter duration, is much more similar to those of SN 1961V and possibly V12 in NGC However, the maximum luminosity of the SN 1997bs outburst is very much like that of the great eruption of η Car. Given the short duration of its eruption, the total energy emitted by SN 1997bs is much less than either η Car or SN 1961V and is probably more like that of V12 or P Cygni (not shown here, but also discussed by Humphreys et al. 1999). SN 1961V and V12 show a flattening (plateau) one to two years past maximum, similar to the behavior shown by SN 1997bs, and years later faded considerably, probably due to the formation of dust in the ejected material (Goodrich et al. 1989; Filippenko et al. 1995). From a color standpoint, LBVs, including η Car, have the spectra of F-type supergiants during eruption (Humphreys et al. 1999), with intrinsic B V 0.3 mag. If this were the intrinsic color of SN 1997bs at outburst, the reddening for SN 1997bs would be E(B V ) 0.37 mag, which is quite close to our estimate of the reddening based on the Na I D absorption line in the optical spectrum. What was the star that gave rise to SN 1997bs? With M V 8.1 mag, it is too luminous to have been an O- type star and was most likely a supergiant later in spectral type than O9 or B0. If the pre-outburst star was in an LBV quiescent state, then it was most likely a mid B-type star, with a surface temperature of 20,000-25,000 KandM bol 10 to 10.5 mag, which is typical for these objects, with initial mass 60 M and a main sequence lifetime of 4 Myr (Humphreys & Davidson 1994; Humphreys et al. 1999; Stothers & Chin 1996). We can attempt to estimate the mass of the progenitor star from an environmental analysis of the properties of other stars in its vicinity, much the same way as in Van Dyk et al. (1999). We have more thoroughly analyzed the HST data containing SN 1997bs by producing very deep images in both the F555W and F814W bands through coaddition of the individual images taken at the various epochs. The results are a s total exposure in F555W and a s exposure in F814W. We show the F555W total image in Figure 7. For the distance modulus µ =30.28 mag measured by Saha et al. (1999), 1 =55 pc. For the WFPC2 chip, then, one pixel 6pc. Wenote that at this resolution, it is difficult to distinguish a single object from a blend of several contiguous ones, so it is possible that the SN progenitor could have exploded among a small, compact star cluster, and that now we see the surviving cluster minus the progenitor. With the available data, we cannot rule out this possibility.

5 5 We have analyzed the environment of SN 1997bs by measuring the magnitudes, using the PSF-fitting techniques described in 2.3, of the stars in a 20 (northsouth) 12 (east-west) region centered on SN 1997bs. Figure 8 shows the resulting color-magnitude diagram. Overlaid are the theoretical isochrones for solar metallicity from Bertelli et al. (1994), after conversion, as in Van Dyk et al. (1999), from the Johnson-Cousins system to the WFPC2 synthetic magnitude system. We also converted the reddening and extinction to the WFPC2 synthetic system. Although many young (massive) dwarf and supergiant stars are resolved in this larger area, and the abundance of dust and luminous stars in the environment implies recent massive star formation, the stars within 1 ( 55 pc) of SN 1997bs are older, less luminous giant and supergiant stars. Regardless of whether the progenitor of SN 1997bs has survived the outburst, the star (or possible surviving compact star cluster) does not seem to have formed in an H II region or an association of obvious O- type stars, unlike the case of η Car, which formed among clusters of young massive stars of similar age (Davidson & Humphreys 1997). However, P Cygni, which may be similar to the progenitor of SN 1997bs, is in the Cyg OB1 association, a group of mostly B-type stars. Based on the diagram in Figure 8, we are unable to place a meaningful constraint on the mass of the SN 1997bs progenitor. To perform a more complete assessment of the massive-star population in the vicinity of SN 1997bs, and to determine whether the progenitor alone survived, or is among a surviving star cluster, higher-resolution U and B images must be obtained, since, in particular, V I is a poor discriminant of hot stars. These blue images would also provide a better estimate of the extinction toward the SN. Two more recent examples of objects similar to SN 1997bs possibly include SN 1999bw in NGC 3198 (Filippenko, Li, & Modjaz 1999) and SN 2000ch in NGC 3432 (Filippenko 2000b). SN 1999bw appears to have similar spectral characteristics to SN 1997bs and is also quite subluminous. SN 2000ch is similar, but may have experienced a shorter-lived outburst than did SN 1997bs (Hudec et al. 2000). The SN IIn subclass clearly spans a very broad range of properties, as was already known to some extent (e.g., Filippenko 1997). If the conclusions of this paper are correct, and the progenitor of SN 1997bs survived, then the explosion mechanism is not core collapse in all SNe IIn. That is, the subclass appears to include a number of supernova impostors such as SN 1997bs, which are evolved but still living. For SN 1997bs, a more definitive test will be whether the star remains visible in future HST images obtained years after the outburst (although, even if it survived, it might fade at optical wavelengths, due to the formation of dust in the ejecta; see Goodrich et al. 1989). The occurrence of SN 1997bs and its possible cousins (SNe 1961V, 1954J, 1999bw, and 2000ch), and their resemblance to η Car and LBV super-outbursts, begs the question of the ultimate fate of these very massive stars. The work of A.V.F. s group at UC Berkeley is supported by NSF grants AST and AST , as well as by NASA grant AR from the Space Telescope Science Institute, which is operated by AURA, Inc., under NASA contract NAS KAIT was made possible by generous donations from Sun Microsystems, Inc., Photometrics, Ltd., the Hewlett-Packard Company, Auto- Scope Corporation, the National Science Foundation, Lick Observatory, the University of California, and the Sylvia and Jim Katzman Foundation. We thank A. J. Barth, D. C. Leonard, and A. M. Gilbert for assistance with the Lick 3-m observations and reductions. We are also grateful to L. A. Wells for providing template CCD images of NGC REFERENCES Bertelli, G., Bressan, A., Chiosi, C., Fagotto, F., & Nasi, E. 1994, A&AS, 106, 275 Bertola, F. 1964, Ann. d Ap., 27, 319 Cavagna, M. 1997, IAU Circ Cardelli, J. A., Clayton, G. C., & Mathis, J. S. 1989, ApJ, 345, 245 Ciatti, F., & Rosino, L. 1977, A&A, 56, 59 Davidson, K., & Humphreys, R. M. 1997, ARA&A, 35, 1 Doggett, J. B., & Branch, D. 1985, AJ, 90, 2303 Figer, D. F., Najarro, F., Morris, M., McLean, I. S., Geballe, T. R., Ghez, A. M., & Langer, N. 1998, ApJ, 506, 384 Figer, D. F., Morris, M., Geballe, T. R., Rich, R. M., Serabyn, E., McLean, I. S., Puetter, R. C., & Yahil, A. 1999, ApJ, 525, 759 Filippenko, A. V. 1991, in SN 1987A and Other Supernovae, ed. I. J. Danziger& K. Kjär (Garching: ESO), 342 Filippenko, A. V. 1997, ARA&A, 35, 309 Filippenko, A. V. 2000a, in Cosmic Explosions, ed. S. S. Holt & W. W. Zhang (New York: American Institute of Physics), 123 Filippenko, A. V. 2000b, IAU Circ Filippenko, A. V., Barth, A. J., Bower, G. C., Ho, L. C., Stringfellow, G. S., Goodrich, R. W., & Porter, A. C. 1995, AJ, 110, 2261 [Erratum: 1996, AJ, 112, 806] Filippenko, A. V., Barth, A. J., & Gilbert, A. M. 1997, IAU Circ Filippenko, A. V. Li, W. D., & Modjaz, M. 1999, IAU Circ Filippenko, A. V., Porter, A. C., & Sargent, W. L. W. 1990, AJ, 100, 1575 Filippenko, A. V., Porter, A. C., Sargent, W. L. W., & Schneider, D. P. 1986, AJ, 92, 1341 Filippenko, A. V., et al. 2000, PASP, in preparation Goodrich, R. W., Stringfellow, G. S., Penrod, G. D., & Filippenko, A. V. 1989, ApJ, 342, 908 Ho, W. C. G., Van Dyk, S. D., Peng, C. Y., Filippenko, A. V., Leonard, D. C., Matheson, T., & Treffers, R. R. 2000, PASP, submitted Holtzman, J. A., Burrows, C. J., Casertano, S., Hester, J. J., Trauger, J. T., Watson, A. M., & Worthey, G. 1995, PASP, 107, 1065 Hudec, R., et al. 2000, GCN Circ. 666 Humphreys, R., & Davidson K. 1994, PASP, 106, 1025 Humphreys, R. M., Davidson, K., & Smith, N. 1999, PASP, 111, 1124 Jeffery, D. J., Branch, D., Filippenko, A. V., & Nomoto, K. 1991, ApJ, 377, L89 Klemola, A. R. 1986, PASP, 98, 464 Krist, J. 1995, Tiny Tim User s Manual, version 3.0a (Baltimore: STScI) Langer, N., Hamann, W.-R., Lennon, M., Najarro, F., Pauldrach, A. W. A., Puls, J. 1994, A&A, 290, 819 MacFadyen, A. I., & Woosley, S. E. 1999, ApJ, 524, 262 Miller, J. S., & Stone, R. P. S. 1993, Lick Obs. Tech. Rep. 66 (Santa Cruz: Lick Obs.) Richmond, M. W., et al. 1995, AJ, 109, 2121 Rupen, M. P., Van Gorkom, J. H., Knapp, G. R., Gunn, J. E., & Schneider, D. P. 1987, AJ, 94, 61 Saha, A., Sandage, A., Tammann, G. A., Labhardt, L., Macchetto, F. D., & Panagia, N. 1999, ApJ, 522, 802 Schaefer, B. E. 1996, ApJ, 404, 411 Schlegel, D. J., Finkbeiner, D. P., & Davis, M. 1998, ApJ, 500, 525 Schlegel, E. M. 1990, MNRAS, 244, 269 Schlegel, E. M., Kirshner, R. P., Huchra, J. P., & Schild, R. E. 1996,

6 6 AJ, 111, 2038 Sollerman, J., Cumming, R., & Lundqvist, P. 1998, ApJ, 493, 933 Stetson, P. B. 1987, PASP, 99, 191 Stothers, R. B., & Chin, C.-W. 1996, ApJ, 468, 842 Treffers, R. R., Peng, C. Y., Filippenko, A. V., & Richmond, M. W. 1997, IAU Circ Turatto, M., Cappellaro, E., Danziger, I. J., Benetti, S., Gouiffes, C., & Della Valle, M. 1993, MNRAS, 262, 128 Van Dyk, S. D., Peng, C. Y., Barth, A. J., & Filippenko, A. V. 1999, AJ, 118, 2331 Weiler, K. W., Panagia, N., & Sramek, R. A. 1990, ApJ, 364, 611 Wells, L. A., et al. 1994, AJ, 108, 2233 Woosley, S. E., Langer, N., & Weaver, T. A. 1993, ApJ, 411, 823 Zwicky, F. 1964, ApJ, 139, 514 Zwicky, F. 1965, in Stars and Stellar Systems, Vol. 8, Stellar Structure, ed. L. H. Aller & D. B. McLaughlin (Chicago: University of Chicago Press), p. 367.

7 7 Table 1 Comparison Star Magnitudes and Colors Star a V B V V R R I a The star numbers are those designated by Wells et al. (1994).

8 8 Table 2 KAIT Photometry of SN 1997bs UT Date JD B V R I 1997 Apr > Apr (03) 1997 Apr (11) 17.12(04) 16.85(02) 16.45(03) 1997 Apr (12) 17.22(07) 16.83(03) 16.48(04) 1997 Apr (05) 17.18(03) 16.87(02) 16.46(02) 1997 Apr (04) 16.97(02) 1997 Apr (07) 17.43(06) 17.01(03) 1997 May (10) 17.40(04) 17.06(02) 16.57(05) 1997 May (17) 17.21(08) 16.73(05) 1997 May (18) 17.61(07) 17.16(04) 1997 May (16) 17.53(11) 17.30(03) 1997 May (04) 1997 May (08) 17.85(05) 17.59(02) 17.16(04) 1997 May (20) 18.12(06) 17.81(03) 17.28(04) 1997 May (06) 18.25(05) 17.89(02) 17.37(04) 1997 May (15) 18.31(06) 17.93(03) 17.46(04) 1997 May (21) 18.24(08) 17.89(04) 17.39(04) 1997 May (12) 17.84(04) 17.21(05) 1997 May (14) 17.41(06) 1997 May (08) 17.86(03) 17.27(04) 1997 Jun (03) 1997 Jun (11) 17.25(03) 1997 Jun (03) 1997 Jun (12) 1997 Jun (12) 18.07(05) Note: The values given in parentheses are the uncertainties in the last two digits of the magnitudes.

9 9 Table 3 HST Photometry of SN 1997bs UT Date JD m F555W m F814W 1997 Nov (03) < Nov (03) 1997 Nov (03) 1997 Nov (03) 1997 Dec (04) 19.49(02) 1997 Dec (05) 1997 Dec (05) 1997 Dec (05) 19.94(02) 1997 Dec (06) 1997 Dec (05) 20.09(02) 1998 Jan (06) 1998 Jan (05) 20.35(05) Note: The values given in parentheses are the uncertainties in the last two digits of the magnitudes.

10 10 Fig. 1. R-band discovery image of SN 1997bs taken on 1997 Apr 15 with the Katzman Automatic Imaging Telescope (KAIT), a 0.75-m fully robotic reflector at Lick Observatory. SN 1997bs was the first supernova discovered by KAIT. The arrow points to the SN. Stars 3 and 4 from Wells et al. (1994) and Table 1 are also indicated. Fig. 2. Spectrum of SN 1997bs obtained with the Lick 3.0-m Shane reflector on 1997 April 16. The absolute flux scale is approximate.

11 Fig. 3. BV RI light curves for SN 1997bs, obtained with KAIT at Lick Observatory, with the addition of the F555W ( V ) and F814W ( I) magnitudes obtained from archival HST images. 11

12 12 Fig. 4. The color evolution of SN 1997bs. Shown are the B V (triangles), V R (squares), R I (pentagons), and V I (stars) colors.

13 Fig. 5. A comparison of the absolute V light curves of SN 1997bs (based on the KAIT and HST Cepheid observations; pentagons) with those of three other Type IIn SNe: SN 1994Y (squares; Ho et al. 2000); SN 1988Z (triangles; Turatto et al. 1994); and SN 1987B (stars; Schlegel et al. 1996). For SN 1997bs, we assume a distance modulus µ =30.28 mag (Saha et al. 1999) and have dereddened the light curve by E(B V )=0.21 mag. Note the relatively low luminosity of SN 1997bs. 13

14 14 Fig. 6. A comparison of the light curve from Figure 5 of SN 1997bs with those of the η Car-like variables, SN 1961V, SN 1954J (V12 in NGC 2403), and η Car. We have adapted schematic light curves for the latter three objects, uncorrected for extinction, from Humphreys et al. (1999, their Figure 1) and the original sources for the photometry. Absolute V light curves are shown for SN 1997bs, η Car, and SN 1954J (Tammann & Sandage 1968); the light curve for SN 1961V is photographic. For SN 1997bs we also show photometry for the progenitor star from an HST archival WFPC2 F606W image made on 1994 December 28 (we have added 1000 days to this epoch in the figure). All of the curves have been adjusted, so that the maxima occur at the same time.

15 15 Fig. 7. The environment of SN 1997bs (indicated by the arrow near the center) on a coadded HST archival WFPC2 F555W image. Fig. 8. Color-magnitude diagram of stars in the environment of SN 1997bs, based on photometry of coadded HST archival WFPC2 F555W ( V ) and F814W ( I) images. The environment represented covers 20 (north-south) 12 (east-west) centered on the SN. Also shown are isochrones from Bertelli et al. (1994) with solar metallicity for ages 4 Myr, 8 Myr, and 30 Myr. The colors of the bluest stars are consistent with the Galactic foreground reddening, E(B V )=0.03 mag. The SN is the (time-averaged) point (filled square) at F555W 23.1 and F555W F814W 3 mag. The other bright red object (triangle), at F555W 22.3 and F555W F814W 2.8 mag, is too bright to be a single star and is probably a red, or reddened, compact star cluster.